摘要
提出了一种基于支持向量回归的点云曲面重构方法,从点云中按一定规则取样得到小样本集,以小样本集为支持向量,并以径向基函数为核函数重建复杂线性函数曲面模型。实验表明该方法能直接重建散乱点云数据,拟合出曲面模型且具有较好的效果,并具有误差小、速度快等优点。
An efficient method for point data sets based on Support Vector Regression is presented. A small sample set is acquired from the point clouds according to the rules, the small sample set is used as suplport vectors. Radial basis function is used as the kernel function of the SVR for modeling linear function surface. Through the experiment, the surface of point sets could be reconstructed directly with the benefits of excellent error accuracy, high efficiency.
出处
《机械》
2009年第3期28-30,共3页
Machinery
基金
国家自然科学基金资助项目(10576027)
关键词
支持向量机
回归机
点云
拟合
support vector machine (SVM)
regression machine
point sets
surface reconstruction